期刊
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
卷 70, 期 3, 页码 2138-2152出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TVT.2021.3060761
关键词
Transportation; Smart phones; Feature extraction; Discrete wavelet transforms; Convolution; Acceleration; Correlation coefficient; Mobile sensing; transportation mode; NIN; Dilate Convolution; GCN
资金
- National Key Research and Development Program [2019YFC1511400]
- Action Plan Project of the Beijing University of Posts and Telecommunications - Fundamental Research Funds for the Central Universities [2019XD-A06]
- National Natural Science Foundation of China [61872046]
- Joint Research Fund for Beijing Natural Science Foundation [L192004]
- Haidian Original Innovation [L192004]
- Key Research and Development Project from Hebei Province [19210404D, 20313701D]
- Science and Technology Plan Project of InnerMongolia Autonomous Regio [2019GG328]
- BUPT Excellent Ph.D. Students Foundation [CX2020221]
- Open Project of theBeijingKey Laboratory of Mobile Computing and Pervasive Device
This paper proposes a novel fusion framework for fine-grained transportation mode recognition, incorporating NIN, Dilate Convolution, and GCN, which outperforms other baseline methods with over 22.3% higher accuracy according to extensive experimental results on the SHL dataset.
Transportation mode recognition is a crucial task of Intelligent Transportation Systems (ITS) in smart city. Though many works have been investigated on transportation mode recognition in recent years, the accuracy and generality are still not able to meet the application requirements. In this paper, we propose a novel fusion framework for fine-grained transportation mode recognition, which consists of the Network in Network (NIN), Dilate Convolution and the Graph Convolutional Networks (GCN). In this framework, we first use NIN and Dilate Convolution to capture local and global features, respectively, and then introduce the graph convolutional network to learn the correlation of features. We construct a topological structure of the features based on the maximal information coefficient (MIC) criteria which is used to measure the similarity between two variables, and then obtain the adjacency matrix used for graph convolution. Extensive experimental results on the public Sussex-Huawei Locomotion-Transportation (SHL) dataset demonstrate the superiority of our proposed NDGCN to other state-of-the-art baselines with more than 22.3% higher accuracy.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据